Research conducted under the Sound Ecosystem Assessment (SEA) Program during the mid-1990s indicated that juvenile pink salmon survival in Prince William Sound was positively correlated with the abundance of the large-bodied copepod, Neocalanus, and negatively correlated with the abundance of walleye pollock. From 2000 to 2006, the Prince William Sound Science Center conducted annual monitoring of the spring abundance and distribution of both macrozooplankton and fish predator populations. The monitoring included multiple frequency acoustic systems and zooplankton net tows. The behavior of large fish predators showed two significant trends, a movement from the deep basins into pink salmon nursery areas and a progressive inshore movement during the odd years that corresponded to poor nursery conditions. Both these movements are consistent with prey-switching behavior from large zooplankton to small near-shore fishes, including pink salmon fry. Correlations between pink salmon adult returns and both large copepod and euphausid abundance in nursery years just missed 90% significance after the first four years of monitoring, despite only three degrees of freedom. However, the correlation weakened with the subsequent two years. The dominant factor in returns over the six-year period is a two-year oscillation. Since these fish are primarily of hatchery origin, and annual releases are similar in magnitude, the basis of the recently manifested two-year cycle is puzzling. An examination of historic runs shows an increasing trend of pink salmon abundance with increasing annual variation. It may be that the juvenile fish are stressing the ecosystem production at recent abundance levels. Under these circumstances, the four year period (2000-2003) of alternating high and low zooplankton abundance may have initiated an ecological mechanism, such as cannibalism, that has continued the two-year oscillation even though the original forcing function no longer matches.